Staff ML Engineer at Deliveroo (2022-06 – Present)
- Hands-on tech lead for revamping Deliveroo's delivery fee pricing algorithm to a competitor-aware econometric pricing model.
- Developed the core model in collaboration with the science leadership.
- Led the MLOps design & implementation for the new training and release pipelines.
- Scoped and prioritised work for a team of 4 ML engineers in the team.
- Impact: Generated £9M incremental revenue in 2023.
ML Tech Lead at J.P. Morgan (2021-08 – 2022-06)
- Developed email classification and named entity extraction models for the companywide email virtualisation platform.
- Moved legacy models and pipelines from on-premise infra to AWS.
- Impact: The system answers ~ 1.5 million queries/month generating additional capacity worth ~ 15 full-time employees.
Co-founder, Senior ML Engineer at emotif.ai (2019-10 – 2021-08)
- Developed an LLM (Bert) based text generation model to generate style-controlled product copy for e-commerce.
- Developed a Shopify app with ML backend and React front end deployed on AWS.
- Open source contributions to tensorflow and truecaser libraries.
- Impact: Reduced copywriting cycle time from days to minutes.
VP, Quant. Researcher at Winton Group (2014-12 – 2019-10)
- Developed attention-based deep learning models for identifying & generating metaphors for style-controlled text generation. This research was spun off as emotif.ai by Winton with £750K in seed funding.